Merge branch 'develop'

Fix missing whitespaces in <li> node

markdownify/__init__.py

@@ -85,9 +85,17 @@ class MarkdownConverter(object):
             convert_children_as_inline = True
 
         # Remove whitespace-only textnodes in lists
-        if node.name in ['ol', 'ul', 'li']:
+        def is_list_node(el):
+            return el and el.name in ['ol', 'ul', 'li']
+
+        if is_list_node(node):
             for el in node.children:
-                if isinstance(el, NavigableString) and six.text_type(el).strip() == '':
+                # Only extract (remove) whitespace-only text node if any of the conditions is true:
+                # - el is the first element in its parent
+                # - el is the last element in its parent
+                # - el is adjacent to an list node
+                can_extract = not el.previous_sibling or not el.next_sibling or is_list_node(el.previous_sibling) or is_list_node(el.next_sibling)
+                if isinstance(el, NavigableString) and six.text_type(el).strip() == '' and can_extract:
                     el.extract()
 
         # Convert the children first
@@ -108,7 +116,10 @@ class MarkdownConverter(object):
 
     def process_text(self, el):
         text = six.text_type(el)
-        if el.parent.name == 'li':
+        # remove trailing whitespaces if any of the following condition is true:
+        # - current text node is the last node in li
+        # - current text node is followed by an embedded list
+        if el.parent.name == 'li' and (not el.next_sibling or el.next_sibling.name in ['ul', 'ol']):
             return escape(all_whitespace_re.sub(' ', text or '')).rstrip()
         return escape(whitespace_re.sub(' ', text or ''))
 

setup.py

@@ -10,7 +10,7 @@ read = lambda filepath: codecs.open(filepath, 'r', 'utf-8').read()
 pkgmeta = {
     '__title__': 'markdownify',
     '__author__': 'Matthew Tretter',
-    '__version__': '0.7.2',
+    '__version__': '0.7.3',
 }
 
 

tests/test_conversions.py

@@ -308,6 +308,10 @@ def test_bullets():
     assert md(nested_uls, bullets='-') == '\n- 1\n\t- a\n\t\t- I\n\t\t- II\n\t\t- III\n\t- b\n\t- c\n- 2\n- 3\n'
 
 
+def test_li_text():
+    assert md('<ul><li>foo <a href="#">bar</a></li><li>foo bar  </li><li>foo <b>bar</b>   <i>space</i>.</ul>') == '* foo [bar](#)\n* foo bar\n* foo **bar** *space*.\n'
+
+
 def test_img():
     assert md('<img src="/path/to/img.jpg" alt="Alt text" title="Optional title" />') == '![Alt text](/path/to/img.jpg "Optional title")'
     assert md('<img src="/path/to/img.jpg" alt="Alt text" />') == '![Alt text](/path/to/img.jpg)'